6 4 BULLETIN OF THE LABORATORIES 



have made upon the cerebellum it may be proper to enumerate some 

 of the results of the late investigations of Golgi (Sulla fma anatomia 

 degli organi centrali de: sistema nervosa, 1886,) of Ramon y Cajal 

 (Revista trimestrial de Histologia normal y patologica, Aug., 1888,) 

 and, more recently, of Prof. Koelliker, which seem not to have attract- 

 ed the attention they deserve in America 



It remains to be seen how far the metallic impregnation upon 

 which Golgi's method rests, develops normal structures and how far it 

 may be relied on to differentiate solely nervous tissues. Really suc- 

 cessful ha^matoxylin and cochineal stains will reveal most of the de- 

 tails, though less conspicuously, and are not ojjen to the objection that 

 the stain may be .a more or less mechanical precipitate rather than a 

 truly selective stain. 



Golgi's chief discoveries are the following: The Purkinje's cells 

 have two sorts of processes, the peripheral protoplasmic processes, 

 which subdivide a'most indefinitely, but do not terminate in , nerve 

 fibres or a nervous reticulum, and a median a.xis-cylinder process, which 

 is occasionally furnished with lateral branches. 



In spite of the numerous subdivisions of the process, anastamosis 

 is said never to take place between different cells. The Purkinje's 

 cells lie between the outer or molecular zone and the deeper or granu- 

 lar zone. 



In the former layer Golgi recognized small cells of variable form 

 with l)oth branching and axis-cylinder processes, which latter also sub- 

 divide but do not reveal their ultimate course. 



In the granular layer Golgi found small nerve cells with a delicate 

 axis-cylinder process and short and few protoplasmic processes 

 extending to granular aggregates. 1'here are also, still more rarely, 

 larger cells of a fusiform shape and very numerous branches of the 

 axis-cylinder. 



C}olgi claims that the medullated nerve fibers in the white layer, 

 and especially in the granular and molecular layers, anastamose largely. 

 This latter fact is not siil)stantiated by Cajal, who, moreover, consid- 

 ers the axis-cylinder of the granular cells, and of the small cells of the 

 molecular layer as well as those of the larger cells of the granular layer 

 as non-medullated. 



Koelliker summarizes his own observations in the following aphor- 

 isms : 



"All t'ree, non-medullated i)rocesses of nerve-fibres are, in my 



